CIMS下基于成组单元的无能力约束的生产批量计划的新方法

The New Method to the Uncapacitated Lot Sizingfor a GT Cell In CIMS

CIMS中基于成组单元的生产批量计划问题是确定属于M个族的N种不同的项目在给定的计划范围T内的每一个时间段上的批量，使得在T内项目总的调整费用（族调整费用和项目调整费用之和）和库存保管费用以及生产费用之和最小（GTLS）本文基于GTLS问题的性质，从一个新的角度即从调整变量出发，运用遗传算法（GA）随机搜索进行求解．对GTLS构造了两阶段启发式算法（Heuristic），通过仿真实验，测试6个问题表明，GTLS／GA比GTLS／Heuristic平均改善5％以内．

The lot-sizing problem in CIMS/GT cell is to determine production lot sizes of N itemgrouped M families over periods that minimizes the sum of setup costs and inventory holdingcosts over the planning horizon, while satisfying given demands. From new view of point,e.g. using setup variables to making chromosomes, we applied genetic algorithm to makestochastic search to solve GTLS based on properties of the solutions of GTLS. A two-stageheuristic algorithm of GTLS is constructed, Computational results of 6 tested problemsshow that the average improvement percent given by GTLS/GA algorithm are within 5%as compared to GTLS/Heuristic solutions.